|Starks, Patrick - Pat|
Submitted to: Agronomy Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 11/3/1996
Publication Date: N/A
Citation: N/A Interpretive Summary:
Technical Abstract: Leaf area index (LAI) is a basic biophysical property used in agricultural and other applications. In the absence of a proven physically-based approach we investigate the use of neural networks to estimate LAI from remotely sensed data. Two line transects were established in a winter wheat (Triticum aestivum L.) field and an alfalfa (Medicago sativa L.) field over seeded with winter wheat. Leaf area index was measured at 3.8 intervals along these transect lines using a LiCor 2000 canopy analyzer. Multi-band remotely sensed data were obtained from the two transects with a NS001 scanner (Thematic Mapper Simulator) flown aboard NASA's C-130. A portion of the remotely sensed and LAI data were randomly selected to train the neural network. The rest of the data were used to test the efficiency of the network to estimate LAI. Two multi-regression statistical techniques were also evaluated. The neural network LAI estimates agreed well with the measured values, and the root mean square error of these estimates was a factor of two lower than those for the statistical techniques for both fields.